Economists and the Analysis of Government Failure: How Cambridge Did and Did
Not Anticipate Chicago and Virginia
Roger E. Backhouse
University of Birmingham
Steven G. Medema
University of Colorado Denver
December 15, 2009
Correspondence: Professor Steven G. Medema, Department of Economics, University of
Colorado Denver, CB 181, PO Box 173364, Denver, CO 80217-3364, USA (email:
[email protected]). We would like to thank Denis O’Brien, Tiziano
Raffaelli, Warren J. Samuels, Richard Wagner, and Donald Winch for their invaluable
advice on previous drafts of this paper. The usual caveat applies.
Economists and the Analysis of Government Failure: How Cambridge Did and Did
Not Anticipate Chicago and Virginia
I. Introduction
The year 2010 marks the 50th anniversary of the publicatoin of Ronald Coase's classic
essay, “The Problem of Social Cost” (1960). Coase's goal in this article, and in
subsequent works such as “The Lighthouse in Economics” (1974), was to point out what
he considered to be the fundamental flaws in the neoclassical theory of market failure.
These flaws took several forms, but central to Coase's critique was the idea that
economists neglected to seriously consider the ability of government to carry out the
policies of market failure correction contemplated by the economists’ theories, a neglect
that was fundamental to any analysis of regulation. Coase summed up these sentiments in
his assertion that in economic analysis “we find a category ‘market failure’ but no
category ‘government failure’” (1964, p. 195). Significantly, Coase laid much of the fault
for this at the feet of A.C. Pigou and the larger Cambridge welfare tradition, the tradition
in which neoclassical welfare economics—including the theories of externalities and
public goods—was grounded (Coase 1960, pp. 28-42; Coase 1974, pp. 357-60, 372-76).
“The Problem of Social Cost” was an early salvo in the “government failure”
literature, fundamental to much work on the economic analysis of law, that developed in
part as a response to or reaction against a perceived shortcoming in welfare economics:
the absence of a theory of governmental behavior that would work in tandem with the
extant theories of consumer, producer, and market behavior. 1 The lack of such a theory, it
was argued, led economists to commit what Harold Demsetz (1969) labeled the “nirvana
The exceptions generally noted here are the work of Knut Wicksell (1896) and the Italian public finance
tradition, which embodied models of government behavior, both efficiency-oriented and “failure”-oriented.
On the Italian school, see, for example, Buchanan (1960) and Medema (2005).
1
2
fallacy”—comparing existing (imperfect) institutions with what an idealized government
could achieve and from this deducing that government can perfectly resolve market
imperfections. James Buchanan saw this problem as pervasive and, like Coase, linked it
to the Cambridge school: “Since Sidgwick and Marshall, and notably since Pigou’s The
Economics of Welfare,” he said, “economists have accepted the presence or absence of
external effects in production and consumption as a primary criterion of economic
efficiency” and suggested that these inefficiencies “can be reduced or eliminated by the
shift of an activity from market to political organization” (1962, p. 17). Buchanan went
on to argue that the various governmental policy measures advocated by economists are
“[i]n almost every case … assumed to be within the realm of the genuinely attainable”
and that objections to this on political grounds are met with the response that the
economist’s task “is not that of the politician, that he does not appropriately concern
himself with the political feasibility or workability of his proposals” (1962, p. 18).
It is no accident that both Buchanan and Coase refer back to the Cambridge
school, given its central place in the development of neoclassical welfare theory. Indeed,
Cambridge has long been portrayed as the interventionist counterpart to the more marketaffirming Chicago and Virginia schools. While there is indeed plenty of evidence that
neoclassical welfare theory was often guilty of the nirvana fallacy, 2 there is good reason
to question the accuracy of applying these sentiments, as Coase and Buchanan did, to the
originators of the Cambridge tradition—Henry Sidgwick, Alfred Marshall, and A.C.
Pigou—and their views of the political process. 3 What we will argue here is that the
See, for example, the surveys of externality theory in Mishan (1971) and of externality and public goods
theory in Cornes and Sandler (1996).
3
We leave aside Keynes because, though he was in many ways a true product of the Cambridge tradition,
sharing many of the values of those considered here, our focus is on market failure and welfare theory, and
Keynes's inclusion thus would raise a host of additional issues, consideration of which would detract from
the main thesis of the paper.
2
3
central texts of the Cambridge tradition evidence a clear sense of the potential limitations
and inefficiencies of the political process that were later developed, albeit in more
systematic fashion, in the government failure literature, and in law and economics and
public choice analysis in particular. In doing so, we will show how the criticisms of the
Cambridge approach for failing to deal with the political process are misguided and at the
same time show how and why the Cambridge view diverges from that associated with the
Chicago and Virginia schools, in spite of their strong similarities.
The issues raised here go beyond the history of economic analysis. The current
economic situation has led economists and policymakers alike to question the ability of
markets to satisfactorily coordinate self-interested behavior in a variety of contexts, and
with this have come calls for increasing government oversight of the market and a
revisiting of the approach to public economics and welfare theory that dominated
economics for much of the twentieth century. What is often missing from these
discussions, however, is the concerns about state action that were central to the early
working out of neoclassical welfare analysis—concerns that were subsequently lost in the
elaboration of the post-war theory of market failure.
II. Sidgwick's “Choice of Evils”
Henry Sidgwick stands at the headwaters of the Cambridge welfare tradition. He took
John Stuart Mill’s classical utilitarian framework for the analysis of market failures and
gave it a marginalist cast while at the same time offering a different view of the role of
the state in response to situations of market failure. 4
A comparison of Mill and Sidgwick's respective treatments of market failure and the economic role of
government can be found in Medema (2007a); see also Medema (2009a, chapter 2). Sidgwick also applied
the utilitarian calculus to crime and punishment and so is something of a bridge between the work of
Bentham and that of Gary Becker. See Medema (2007b).
4
4
In his essay on “Economic Socialism” (1886), Sidgwick identified two basic
categories of market failure: those instances where the pursuit of self-interest does not
generate the wealth-maximizing result, and those situations in which self-interest does
generate the wealth-maximizing result but in which that result is inconsistent with the
larger social interest.5 For Sidgwick, however, the fact that market outcomes are not
always wealth-maximizing or otherwise in society’s best interests did not in itself mean
that government intervention is necessarily the best course of action. “[I]n human
affairs,” he said, “we have often only a choice of evils,” and while private activity may
yield less than satisfactory results at times, “it is possible that governmental interference
might on the whole make matters worse” (1886, p. 206). Sidgwick expressed a similar
sentiment in his Principles of Political Economy: “It does not of course follow that
wherever laisser-faire falls short governmental interference is expedient; since the
inevitable drawbacks and disadvantages of the latter may, in any particular case, be worse
than the shortcomings of private enterprise” (1901, p. 414). 6
In support of this view, Sidgwick pointed to several “drawbacks and
disadvantages” associated with government intervention. 7 First, the expanding scope of
governmental operations adds to the patronage possibilities at the disposal of government
agents, and this, in turn, raises “the danger of increasing the power and influence capable
of being used by government for corrupt purposes” (1901, p. 414).
Second, legislative actions are heavily influenced by the desire to please special
interests groups, which leads to “the danger … that the exercise of [the legislature's]
economic functions will be hampered and perverted by the desire to gratify influential
Elsewhere, Sidgwick (1901, Book III) elaborated an extensive list of such market failures.
The first edition of Sidgwick's Principles was published in 1883.
7
See also Sidgwick (1897, pp. 167-68).
5
6
5
sections of the community—certain manufacturers, certain landlords, certain classes of
manual laborers, or the inhabitants of certain localities” (1901, p. 414).
Third, Sidgwick questioned the ability and willingness of legislators to govern in
the national interest, particularly in light of his belief that politicians take positions to
attract votes, rather than seeking votes based on exogenously determined positions.
Sidgwick’s discussion brings to mind the concept of the vote-maximizing politician and
the pork barrel process when he says that
when legislation is in the hands of a representative assembly, the more the
functions of government are extended in a socialistic direction, the greater
becomes the risk that contested elections will exhibit an immoral competition
between candidates promising to procure public money for the benefit of
particular classes and districts (1897, p. 167).
Note that Sidgwick was also making what amounts to the Leviathan argument here—that
the growth of the state feeds on itself to promote more growth via the competition for
votes in the political process.
Sidgwick was also concerned about the influence of imperfect information on the
political process. He was critical of what he called “wasteful expenditures” made “under
the influence of popular sentiment.” 8 Here, he argued, legislators can exploit voter
myopia in order to garner support from the electorate at low cost. The basic problem, said
Sidgwick, is one of information and perspective: the general population, “however
impatient of taxation, are liable to be insufficiently conscious of the importance of thrift
in all the details of national expenditure” (1901, p. 414). He also said that people are
much more likely to know what is in their best interests, and to undertake courses of
Sidgwick did not provide us with a definition of “waste,” a concept which has a fairly precise meaning in
the modern public choice literature. As such, the modern definition may or may not apply here.
8
6
action to facilitate that, than is Government (1897, p. 144). And even as progress in
society was introducing a degree of complexity that could make it more difficult for
individuals to discern what is in their best interests, Sidgwick was not confident that
government could do any better (1901, pp. 416-17).9
While Sidgwick tended to focus on potential problems within the political arena,
he was not unaware of those associated with government bureaucracy. He expressed
particular concern about the inefficiencies that come with the continued expansion of the
range and complexity of government activities, especially on the supervisory front (1901,
p. 414). Sidgwick also pointed to the lack of incentives for government workers to
properly carry out the functions assigned to them: “the work of government,” he said,
“has to be done by persons who—even with the best arrangements for effective
supervision and promotion by merit—can only have a part of the stimulus to energetic
industry that the independent worker feels.” Private sector workers “may reasonably hope
to gain by any well-directed extra exertion, intellectual or muscular, and most fear to lose
by any indolence or neglect,” whereas these stimuli are largely lacking in the public
sphere (1901, p. 415).10
So, while Sidgwick did point to a fairly wide range of phenomena that he
considered market failures, he came down in the end saying that, “on the whole, there
seems no doubt that even where the defects of laisser faire are palpable and grave, they
may still be outweighed by the various disadvantages incident to governmental
management of industry” (1901, p. 416). Reflecting the marginalist framework within
Sidgwick cited as an example here sellers pedaling products that seem useful but really are not, and that
consumers would not tend to purchase if they had good information about the products.
10
Sidgwick went on to say, though, that this also applies to joint stock companies, and he even suggested
that government may have better means than some private concerns to get around these incentive problems,
in that it has the ability to “reward conspicuous merit by honours and distinctions,” as well as offering
greater employment security (1901, p. 415).
9
7
which he was operating, Sidgwick contended that the various “costs” that he has pointed
to must be weighed against the benefits to determine the appropriateness of government
intervention. And, even when intervention is in the public interest, this interference, he
said, should be as mild and as narrowly drawn as possible while still accomplishing its set
goal.11
Although Sidgwick raised many issues central to modern public choice concerns,
he parts company with the modern view when he suggests that there is good reason for
long-run optimism regarding the possibilities of government intervention. Sidgwick
believed that “moral and political progress [in society] may be expected to diminish” the
extent and severity of the shortcomings that he has been discussing (1901, p. 416,
emphasis added), which will increase over time the range of activities that government
can carry out in a manner superior to market forces. While Sidgwick is silent on the
specifics here, the background to this hope was presumably the British experience of
reforming the Civil Service by measures such as recruiting through competitive
examinations. The accuracy of Sidgwick's optimistic prognostication—which, as we shall
see, was later echoed by Marshall and Pigou—is beside the point for present purposes.
What matters is that Sidgwick is here arguing that the appropriate scope for government
action depends on the government’s competence: in short, he is not viewing the
government as an omniscient maximizer of social welfare.
Political considerations are also in the background to Sidgwick’s discussion of redistribution of income.
He offers utilitarian arguments for egalitarian redistribution but draws back, on the grounds that once the
government had embarked on this, it would likely be taken too far (Sidgwick, 1901, p. 565).
11
8
III. Marshall and “the Money Value of Political Power”
In contrast to Sidgwick and Pigou, Marshall did not devote a great deal of attention to the
theory of economic policy. As a consequence, his remarks on the role of the political
process in economic affairs were usually written in response to concrete issues of policy
on which he decided to comment or offer advice. As trade policy and socialism were very
much in the air during Marshall’s professional life, it is perhaps not surprising that a
goodly amount of his commentary on the role of the state takes place against these
backdrops.
Marshall believed that the expansion of industry and trade since the middle ages
had both positive and negative implications for the ability of government to be able to
manage economic affairs in the public interest. On the positive front, given the vast
expansion of economic activity during this period, “The possible functions of
Government expanded almost obtrusively in many directions” (ibid). On the other hand,
we did not see a corresponding expansion of state action because “the great increase in
the complexity of the problems of industry” was accompanied, “for the time at least,” by
only a “small an increase in the mental capacity of Government for dealing with them”
(1923a, p. 42).
A second effect of economic progress, according to Marshall, was that it put the
common people in a position where they were “not very far inferior in shrewdness, and
constructive ability to the ruling classes” (1923a, p. 42). The practical import of this
development was that the advantage the common man had over government officials
because of “the more intimate knowledge possessed by each of his own circumstances,
and the technical problems of his own work” was only increasing. This, said Marshall,
made it likely that industry and trade would develop most advantageously if left to take
9
their own course (1923a, p. 42). This increase in the capabilities of the common people
relative to the ruling class brought about “a slow increase in the power of the people to
govern the Government that governed them” and thus provided some degree of check on
state action (1923a, pp. 42-43).
Like Sidgwick, Marshall claimed that “Almost every extension of Government
activity brings with it good and evil, both economic and political” (1923a, p. 496). What
particularly concerned him were politically motivated proposals for significant changes in
economic policy. These, he said, “are likely to bring about results which will not be
satisfactory from the economic point of view, and will perhaps introduce morbid
elements into politics.” The chief problem that Marshall saw arising from the interplay of
economic and political forces was the incentive for corruption, which was a recurring
theme in his discussion of the political process. For example, in a memorandum prepared
for the Treasury on trade policy in the early 1900s, 12 Marshall said that he had learned
from his 1875 visit to the U.S. that protection was in practice very different from the ideal
described by protectionists. Protective policy became intricate, and “in becoming intricate
it became corrupt, and tended to corrupt general politics” (Marshall 1926, p. 394). 13 He
considered that this “moral harm far outweighed any small net benefit” (ibid). In
England, by way of contrast, Free Trade had reduced “the money value of political
power,” a process that would have been reversed had the movement for Tariff Reform
(raising tariffs so as to be able to offer preferential treatment to the Empire) been
successful (Marshall 1926, p. 395).
The history of this document, the first draft of which was lost in the mail, is discussed by Groenewegen
(1996). It was concerned with issues of imperial preference and the use of tariffs as a weapon in bargaining
with other countries (Groenewegen 1996, p. 300).
13
This is repeated verbatim in Marshall (1923b, pp. 219-220).
12
10
In fact, Marshall seems to have considered corruption endemic to trade policy.
Rent seeking was certainly a culprit, as those who stood to gain from such policies tried
to manipulate policy making to their benefit. These policies also impacted the officials
who administered them by “increas[ing] their temptations to use public authority for the
purposes of private gain” (1923a, p. 43). Marshall believed that tariffs were imposed to
do little more than gratify powerful interest groups (1926, p. 379), and this underlies his
emphasis that protection, once imposed, was difficult to remove (Marshall 1923b, pp.
218ff). Indeed, the historical record convinced Marshall that such problems were nearly
impossible to avoid, and that corruption tends to increase when there is “much money to
be gained by political influence” (1923a, p. 43). So strong were Marshall's feelings on
this score that his policy recommendations in support of free trade rest not on the
theoretical apparatus he built to analyze trade, but on government failure arguments.
Marshall’s discussion of the nationalization of the railways provides an excellent
illustration of his concerns about the problem of rent seeking. 14 Marshall points out that,
under nationalization, rank and file railroad workers will be able to exert so much
pressure through the voting booth that “political considerations” will govern decisions
when those decisions ought to instead be made on “technical grounds.” In such cases, the
“the interests of the population as a whole” will be wrongly sacrificed to the special
interests of one segment of the population (1923a, p. 494). 15 Part of the problem here is
that the political motivations of legislators are an impediment to sound policy, as
legislators are controlled not by their constituents but by powerful interest groups
Marshall’s discussion of the nationalization of railways relies heavily (and approvingly) on a paper read
by Mr. Acworth at the 1912 Congress of the Royal Economic Society.
15
He contends that the owners of the railroads will benefit because “the State seldom buys a property
without paying more for it than it is worth,” and the owners know this. Moreover, railway officials have no
cause for worry because most of them will end up with salaries that are at least as high and positions that
are somewhat more prestigious (1923a, p. 494).
14
11
(Marshall 1926, p. 396).16 However, it is not only the politicians who are to blame,
according to Marshall, nor is he ambiguous about the results: “If particular groups of the
people look on the polls, as opportunities for advancing their own sectional interests,
rather than for rendering an upright and dutiful service to the State, the Mother of all,” he
says, “then the nation as a whole will become less noble, weaker and ultimately poorer
…” (1923a, p. 495).
Like Sidgwick, Marshall also expressed concern about problems raised by the
presence of imperfect information on the policy front—in particular, politicians’ habit of
exploiting informational asymmetries for their own benefit, especially via the masking of
costs. He felt that politicians tend to shade the truth, selecting facts which fit their case
but ignoring those that support the other side, in order to have their way without losing
the support of the electorate. This, according to Marshall, is one of the factors that
distinguishes the politician from the businessman, for whom economic factors are
dominant. For example, there are certain risks which businessmen will take into account,
but which are “apt to be neglected” when political considerations dominate economic
ones (1923a, p. 496).
Marshall also lamented the propensity of legislators to put into place policies
where the costs are spread sufficiently widely so as not to attract notice, even as the
benefits are concentrated among a small group of interested parties:
Unfortunately the experience of many centuries shows that a policy, which
will confer a considerable benefit on each of a compact group of traders or
producers, will often be made to appear to be in the interest of the nation;
because the hurt wrought by it, though very much greater in the aggregate
16
Marshall cites the German Reichstag as an example of this problem.
12
than the gain resulting from it, is so widely diffused that no set of people are
moved to devote much time and energy to making a special study of it. Its
advocates speak with zeal and the authority of expert knowledge. But they are
bad guides, even if unselfish and perfectly upright: for a policy that makes for
their peculiar profit is invested in their eyes with a deceptive glamour.
(Marshall 1917, pp. 329-30)
Policy-related costs, then, are often underestimated by the public because they are not
obvious. Marshall sees this occurring on the trade policy front, where one of several
reasons why inappropriate tariffs would be levied is that it is easy to know who benefits
from a tariff, whereas those who suffer are harder to identify (1926, p. 390). Not
surprisingly, Marshall believes that government is more likely to be “vigilant” when costs
are easily traceable than when they are not (1923a, p. 671). He suggests that State
management has the most to offer in areas consisting of “routine operations” that are
carried out very much in the public view, and where there is little capital expenditure, so
that expenses and actual undertakings correspond very closely—such as with the postal
service (1923a, p. 668).
It is not just politicians/legislators who are the subject of Marshall’s concern; he
also identified several reasons why government bureaucracies produce undesirable
outcomes. The first was that the Minister who is charged with overseeing a particular
bureau, and upon whom final decision-making power rests, is also someone whose
“knowledge of the technique of his Department is seldom so great.” These officials have
many other duties to occupy their minds and need to spend a great deal of effort on
political machinations relating to policies that will maintain or enhance the position of
their ministries. While this work is important for the Minister and for his ministry, says
13
Marshall, “in regard to his responsibilities as head of a vast business it is a grievous, if
not disastrous hindrance” (1923a, p. 666). As a result, these Ministers are more likely
than their private-sector counterparts to rely on the recommendations and advice of their
subordinates. The private sector leaders, in contrast, devote all off their efforts to a more
narrow set of issues concerned with their particular business (1923a, p. 666).
A second bureaucratic problem identified by Marshall is that the rank of the
bureaucrats—the subordinates on whose advice the Ministers rely—is largely determined
by seniority rather than ability. Marshall suggests that the seniority system was put into
place because of fears that political influence would govern promotion absent a codified
system like seniority (1923a, p. 496). Nor was Marshall optimistic that the situation
would change, as, in his estimation, those who benefit from the seniority system see its
advantages for themselves, and they weigh these personal advantages more heavily than
the shortcomings that impact society as a whole (1923a, pp. 666-67).
Third, political control over bureaucracies and the ability to effectively monitor
bureaucratic activities is often weak, and legislators will tend to rubber stamp
bureaucratic decisions (Marshall 1923a, p. 670). And, he said, when politicians did
attempt to exercise oversight over the bureaucracy, their motives—gaining votes from
their constituents—and thus the results, were something less than optimal (Marshall
1890, p. 275) This, according to Marshall, would tend to produce “political immorality”
(ibid, p. 276).
When one combines Marshall’s concerns about the behavior of politicians with
his suspicions about the functioning of the bureaucracy it should come as no surprise that
he believed that Government action was inimical to progress. Indeed, it was private
enterprise, not government, that had been responsible for rising productivity, he said. This
14
was a view that Marshall expressed on multiple occasions and in very strong terms.
Witness this statement, from his 1907 essay on “Economic Chivalry”:
Government creates scarcely anything. If Governmental control had supplanted
that of private enterprise a hundred years ago, there is good reason to suppose that
our methods of manufacture now would be about as effective as they were fifty
years ago, instead of being perhaps four or even six times as efficient as they were
then. (Marshall 1907, p. 338)
For someone attracted to socialism, 17 this was a remarkably blunt and unambiguous
statement. Why Marshall took this position is evident from his commentary on the
possibility of nationalizing the railways, discussed above.
What gives Marshall’s discussion of these issues an air that seems dated in
relation to modern government failure theory is his emphasis on the variability of human
character and the ability of men (he normally meant men, not women) to rise above
narrow considerations of self-interest. Marshall believed that the environment strongly
influenced character, and that some men could be trusted to act in the public interest,
whereas others could not. Marshall saw the pervasiveness of corruption playing a major
role in Adam Smith’s (1776) suspicions regarding state action and spoke directly about
the low moral state of government in Smith’s time. However, said Marshall, Smith “had
no means of anticipating the vast increase in the resources of Government, and in the
honesty of public officials which began in the nineteenth century” (1923a, p. 45). 18
During Mill’s lifetime, “the probity, the strength, the unselfishness and the resources” of
government increased (ibid., p. 335), enabling it to do more. Causes of this included
See, for example, Marshall’s letter to Lord Reay of November 12, 1909 (in Pigou 1925, pp. 461-65).
Marshall goes so far as to suggest that “Adam Smith’s doctrine, carefully interpreted, supports
[government’s] active intervention in many affairs in an age in which it has acquired the power and the will
to govern the people wisely; and the people have acquire the power and the will to govern their
Government with knowledge, discretion and restraint” (1923a, p. 719).
17
18
15
Parliamentary reform (ending what historians know as “the Old Corruption”), the spread
of religious zeal, the improvement and spread of literature, and even the rise of cooperation.
Marshall contended that the ability of government to beneficially intervene
increased at an even more rapid clip after Mill’s death. In addition to the spread of
general education and “a general surplus of energy over that required for earning a
living” (1907, p. 336), Marshall pointed to technological factors such as “shorthand
reporting, the electric telegraph, and the improved printing press,” which, he said, “have
given strength to the general movement towards higher ethical standards.” This, in turn,
“has been steadily cleansing Parliament, and invigorating Governmental departments
(Marshall 1926, p. 395). Good communications and the free flow of information, then,
reduce the potential for corruption.
Thus, said Marshall, we have come to a point at which “the power of the people to
govern the Government that governed them … was destined to become so large and farreaching, that many tasks may now reasonably be entrusted to Government in the
twentieth century which would have been grossly mismanaged in the first half of the
nineteenth: thus a certain new tendency to a widening of the appropriate functions of
Government gradually set in” (1923a, pp. 42-43).
The problem, according to Marshall, was that, though the potential for
government had increased, the complexity of the tasks facing it had increased even
further. Moreover, all of these advances in governance and Marshall’s optimistic view of
the possibilities offered by them do not seem sufficient to overcome what seemed to him
almost inherent problems in the operation of government. In spite of all of these gains, he
says,
16
it seems to remain almost as true now, as in former times, that the heavy hand of
Government tends to slacken progress in whatever matter it touches; and finally
that ‘business influences are apt to corrupt politics; and political influence are apt
to corrupt business.’” (1923a, p. 672)
While some believed that committees made up of experts could rise above these
problems, Marshall argued that even using expert committees does not guarantee the best
results, in that these committees will take outside advice, and this advice often reflects
biases and vested interests (1923a, p. 671).
IV. Pigou and the “Prima Facie Case” for Intervention
Though Sidgwick and Marshall played an important role in developing the analysis of
market failure that became central to neoclassical welfare theory, the analysis reached full
flower in the work of A.C. Pigou, whose Wealth and Welfare (1912) and Economics of
Welfare (1920/1932) provided the foundation upon which the modern theory of market
failure was erected. Pigou's case for government intervention is well known: it was first
laid out in Wealth and Welfare and repeated in his subsequent writings on the subject. It
subsequently became the target of criticism by Coase (1960) and others. The most wellknown statement of the case comes from Economics of Welfare, where Pigou said that
“[i]n any industry, where there is reason to believe that the free play of self-interest will
cause an amount of resources to be invested different from the amount that is required in
the best interest of the national dividend, there is a prima facie case for public
intervention” (1932, p. 331). The question, of course is what this actually implies for
policy.19
19
See also Pigou (1935): When “private self-interest, acting freely, subject only to the ordinary forms of
law, does not lead to the best results from a general social point of view, there is … a prima facie case for
17
The political element weighs heavily for Pigou in determining whether and how
the prima facie case translates into practice. Pigou touched on this briefly in Wealth and
Welfare,20 noting that “The case … cannot become more than a prima facie one, until we
have considered the qualifications, which governmental agencies may be expected to
possess for intervening advantageously in this class of matter” (1912, p. 247). Pigou
sounds a theme later emphasized by Coase21 when he argues that idealistic
characterizations of government are not helpful in evaluating these qualifications:
It is not sufficient to contrast the imperfect adjustments of unfettered private
enterprise with the best adjustments that economists in their studies can
imagine. For we cannot expect that any State authority will attain, or even
whole-heartedly seek, that ideal. (1912, p. 248)
He cited sectional interests, corruption, and politics as factors that would cause policy to
fall short of the ideal given by theory. In an essay on “State Action and Laisser-Faire,”
Pigou elaborated on the implications:
In order to decide whether or not State action is practically desirable, it is not
enough to know that a form and degree of it can be conceived, which, if carried
through effectively, would benefit the community. We have further to inquire how
far, in the particular country in which we are interested and the particular time that
concerns us, the government is qualified to select the right form and degree of
State action and to carry it through effectively. (1935, p. 124)
Ascertaining the qualifications of government, in turn, has several components.
State action” (1935, p. 124).
20
It is repeated almost verbatim in Economics of Welfare (1932).
21
See, for example, Coase (1966, p. 100; 1988, pp. 19-20, 28-30).
18
First, the attitude of the citizens toward government action is of great import, as
any actions are more likely to be effectively implemented and carried out if they have
substantial public support. Second, there is the issue of the “quality” of the decisionmaking body that will be determining the form and extent of state action. At this point
one can hear in his remarks strong echoes of Marshall’s and Sidgwick’s skepticism about
the capabilities of the state to successfully resolve these market failure problems. Pigou
said that the quality of the decision-making body will be a function of:
the intellectual competence of the persons who constitute it, the efficacy of the
organisation through which their decisions are executed, their personal integrity in
the face of bribery and blackmail, their freedom from domination by a privileged
class, their ability to resist the pressure of powerful interests or of uninstructed
opinion (1935, p. 125).
The basic problem, according to Pigou, is that “Every public official is a potential
opportunity for some form of self-interest arrayed against the common interest” (1912, p.
248). The financial stakes that accompany many policy proposals are incredibly high and
make “Logrolling and lobbying” “powerful forces” that “are certain to be called into
play” (1935, p. 126). Because of this, politicians are “subject to great pressure from
persons who can control votes,” a problem that is made all the more difficult by the fact
that the lobbying groups are also the sources of politicians’ campaign funds (1912, p.
248). Moreover, the growth of government enterprises generates substantial new
bureaucracies and “the employment of tens of thousands of additional public servants,”
which, in turn, means increased patronage for party leaders (1912, p. 248). All of this
makes the quality of the individuals who make up the governing body of the utmost
import for sound policy. Unfortunately, says Pigou, the unseemly aspects of interest-
19
group pressure tend to deter upright people from entering politics, which only serves to
further loose the forces of corruption. The results of the political process thus will likely
be skewed, in that the efforts of the state are “most likely to be invoked successfully by
the strong,” regardless of what is actually in the best interests of society (1935, p. 126).
For Pigou, it is practical concerns of a political nature, rather than “any abstract
plea for laisser-faire in matters of trade,” that speak most loudly against government
interference. This is nicely illustrated in his discussion of public versus private
management of industry. Public management, he said, would generally be less efficient
than private management: commercially weak public enterprises would be protected from
competition, perhaps by subsidies; public managers would take too few risks; and public
enterprises will not be the optimal size, for their size will be determined by the size of the
political units running them (1912, pp. 277, 281, 284). These problems with public
enterprise have to be balanced against the inefficiencies likely to arise under monopoly or
monopolistic competition and the difficulty of balancing supply and demand when there
is government regulation of private industry (1912, p. 288).
Pigou followed Marshall in believing that, although government in Adam Smith’s
day was corrupt, inefficient and capable of achieving little, the 150 years since Smith had
seen the rise of “governing authorities … enormously better equipped for successful
action” (p. 126). Referencing Marshall’s “Economic Chivalry” essay, Pigou attributed
this at least in part to increased levels of education and to the “surplus energy” that was
made possible by increases in the technology of production, and he praised both the Civil
Service for its “high capacity and unquestioned public spirit” and the politicians for their
absence of personal corruption (p. 126).22
22
See also Pigou (1912, p. 249).
20
Yet, while the quality of the individuals serving on the political front was
improving, Pigou was not so naïve as to suggest that governing bodies had reached
anything like perfection. In particular, he pointed to four shortcomings that characterize
representative assemblies. First, elected officials are chosen for a host of reasons
unrelated to economic regulation, and there is thus little reason to expect that they have
any particular competence for intervening in industry. Second, the make-up of these
assemblies tends to fluctuate a great deal over time, depriving society of the benefits of
experience and, with that, what may be more desirable levels of policy continuity. Third,
municipal size and appropriate economic size may be vastly different and thus attempting
to set up a public enterprise on the same scale as the municipality may lead to major
inefficiencies. Finally, these assemblies are political in nature and thus subject to
“injurious forms of electoral pressure” of the types we have mentioned above (1912, pp.
249-50).
While the legislative option is problematic, Pigou argued repeatedly that recent
advances in governance structures offer a way around these problems.23 Specifically, he
argued that quasi-governmental entities—he mentioned Public Service Boards and
Commissions—that are not directly subject to political control offer a means of avoiding
some of these problems that was not present in times past. While members of general
governmental bodies are elected for a host of reasons often unrelated to the regulation of
industry, the specialized boards and commissions offer the opportunity to appoint
members whose abilities are well-suited to the regulation of industry. The continuity
problem and the influence of electoral whims can be resolved by appointing these
individuals for longer terms of service than are standard for elected officials. This would
23
See Pigou (1912, p. 250; 1932, pp. 333-35; 1935, p. 127).
21
have the further beneficial effect of insulating them from electoral pressures (1932, p.
333-35; 1935, p. 127). In Wealth and Welfare (1912, p. 249), Pigou said that “This
important fact implies that there is now a greater likelihood of interference, in any given
case by any given governmental authority, proving beneficial than there was in former
times.”
V. Conclusion: After Cambridge, Why Chicago and Virginia?
The government failure movement was born in part as a reaction against neoclassical
welfare economics. The charge was that the neoclassical approach put forth an
interventionist program on both the micro and macro sides while lacking any theory of
the political process. As such, the analysis neglected to account for the difficulties that
attend the policy-making process, and which could render such interventions damaging to
the public interest. What we have shown here is that this caricature is largely erroneous as
regards the Cambridge school, which, partly because of Coase’s critique of Pigou, has
come to be seen as the most significant exemplar of neoclassical welfare economics. 24
From Sidgwick, through Marshall and Pigou, we see neither a neglect of government nor
the portrayal of an omniscient and beneficent government that will deftly right the
perceived wrongs of the market. Rather, the Cambridge school evidenced a profound
awareness of the potential pitfalls of and inefficiencies associated with the political
process, including those related to logrolling, special interests, rent seeking, bureaucratic
behavior, and the like. Where the Cambridge approach parts company with Chicago and
Virginia is in their optimism regarding the improvability of government, grounded in the
potential offered by certain reforms in the institutional structures of political decision
24
The unfamiliarity of this claim is the reason why we have documented our paper so heavily with
quotations.
22
making. There was a strong sense that these reforms had, through the nineteenth century,
facilitated efficiency by mitigating certain of the problems associated with the political
process.25 However, Sidgwick, Marshall, and Pigou never confused improvability with
the possibility of reaching a social optimum.
Interestingly, the Cambridge optimism about the improvability of government has
its parallels in Adam Smith. Smith, of course, was very critical of the politicians of his
day, and many of his criticisms are echoed in both the Cambridge tradition and
contemporary government failure theory. Smith did not, however, see this as an
inevitably permanent condition; rather, he felt that behavior could be improved via
education. As George Stigler has put it, Smith’s “attitude toward political behavior was
not dissimilar to that of a parent toward a child: the child was often mistaken and
sometimes perverse, but normally it would improve in conduct if properly instructed”
(1971b, p. 142). Stigler, who considered the self-interest assumption a hallmark of the
Wealth of Nations, chastised Smith for apparently failing to recognize that political
agents, too, are self-interested in their behavior, and that the link between self-interest
and political decision making is not severed by education. In essence, Stigler was
criticizing Smith for not being a prototype public choice economist. 26
Stigler’s criticism of Smith offers some insight into why the contemporary theory
of government failure does not share the Cambridge school’s optimism regarding the
possibilities for government, even though the spread of education and the various
governance-improving reforms pointed to by Sidgwick, Marshall, and Pigou are very
It bears noting that the rationale behind this sentiment, too, evidences a recognition of the types of
incentive-alignment issues that were later to concern law and economics and public choice scholars.
26
Stigler was not willing to allow that Smith’s failure here could be explained by the fact that everyone
else, too, looked at political behavior in non-self-interested terms; Smith, he says, “is a better man than
everyone else.” (1971b, p. 143), and so should be above such slip-ups. We take no position here about
whether Stigler is correct in his beliefs about the role of self-interest in Smith’s system, but see Medema
(2009b) for a discussion of the heterogeneity of Chicago views of Smith on this score.
25
23
much in evidence today. At the heart of the matter is the current view of agents—
including those operating in the political sphere—as rational maximizers. Rational
maximization is said to pervade political behavior, and this basic motivation is not
affected by education, position, or circumstance. The Cambridge economists, in contrast,
operated with a different view of the individuals involved in government and private
economic institutions. This view is not formalized and hence their results appear to
modern economists to be inadequately theorized. 27 Where the modern theorist sees agents
as rational maximizers, the Cambridge economists, and above all Marshall, saw human
motivations as heterogeneous and formed by the environment in which men found
themselves. What the Cambridge literature essentially posits is politicians who are
subject to pressures because of their place in the electoral process. While not denying that
self-interest is a fundamental human motivation, it was thought that education and moral
values could cause men to take decisions that were in the public interest—though this
process was not completely reliable. We can see this non-homogeneity of motivations in
Pigou’s discussion in Wealth and Welfare. Having laid out his views on the shortcomings
of political agents and processes, he goes on to say that “The force of this argument for
non-interference by public authorities is, clearly, not the same at all times and places, for
any given kind of public authority will vary in efficiency and sense of duty, with the
general tone of the time” (1912, p. 249), It was an elitist view, where educated, Anglo-
27
Marshall, the dominant figure in the Cambridge School, made it clear that the scope for formal theory
was very limited, and that it was unsafe to draw policy conclusions from it. The Cambridge School’s
analysis of the political process and its effects on policy, therefore, did not rest on formal analysis, and they
also did not use the theoretical models and techniques that have allowed public choice economists to reach
more clear cut conclusions—the same models and techniques, ironically, that neoclassical welfare
economists used to “demonstrate” market failure. It is not clear how far this is because they did not have
access to the theoretical techniques on which contemporary government failure results rest, or because they
chose not to make the assumptions that these results would require.
24
Saxon men were seen as able to achieve things that other races and classes were not able
to achieve, a view that is rightly alien to modern economics. 28
The Cambridge optimism about the prospects for state action has more to do,
though, with the perceived insulation of civil servants, or bureaucrats, from electoral
pressures, the use of experts, and the ability of the expert to rise above political concerns.
The economic theory of bureaucracy, of course, suggests that the absence of electoral
pressures is no barrier to inefficiency; expert and non-expert alike respond to incentives
in predictable ways. Likewise, the capture theory of regulation, developed by Stigler,
Peltzman, and others in the 1970s,29 suggests that there is less reason for optimism about
these appointed commissions of specialists than the Cambridge view indicates.
What emerges, then, is that the difference between the Cambridge economists and
their modern counterparts in Chicago or Virginia was not that the former were guilty of
committing the “nirvana fallacy,” or that they were naïve about political processes. The
differences lie in the facts that Chicago and Virginia developed techniques for analyzing
political processes that were not available to the Cambridge economists. The rational
choice approach and its assumption of stable preferences was central here, effectively
ruling out the idea that public interest could supersede self-interest in the calculations of
political actors and that political reforms, education, and the like might be mechanisms
for facilitating this. That is, though Cambridge, Chicago, and Virginia shared an
understanding of the potential for wide-spread government failure, the implications of the
rational choice approach to regulatory and political processes led economists associated
Sandra Peart and David Levy (2005) provide an analysis of these phenomena in nineteenth-century
economics.
29
See, e.g., Stigler (1971a) and Peltzman (1976).
28
25
with the Chicago and Virginia traditions less optimistic that government failures could be
overcome.
References
Buchanan, James M. 1960. “La Scienza delle finanze: The Italian Tradition in Fiscal
Theory.” In James M. Buchanan, Fiscal Theory and Political Economy: Selected
Essays. Chapel Hill, NC: University of North Carolina Press.
______. 1962. “Politics, Policy, and the Pigovian Margins.” Economica 29 (February):
17-28.
Coase, Ronald H. 1960. “The Problem of Social Cost.” Journal of Law and Economics 3
(October): 1-44.
______. 1964. “Discussion.” American Economic Review 54 (May): 192-97.
______. 1966. “The Theory of Public Utility Pricing.” In The Economics of Regulation of
Public Utilities: Papers Presented at a Conference at Northwestern University,
June 19-24, 1966. Evanston, IL: Northwestern University, 96-106.
______. 1974. “The Lighthouse in Economics.” Journal of Law and Economics 17
(October): 357-76.
______. 1988. The Firm, the Market and the Law. Chicago: University of Chicago Press.
Cornes, Richard and Todd Sandler. 1996. The Theory of Externalities, Public Goods, and
Club Goods, 2nd edn. Cambridge: Cambridge University Press.
Demsetz, Harold. 1969. “Information and Efficiency: Another Viewpoint.” Journal of
Law and Economics 10 (April): 1-21.
Groenewegen, P., ed. 1996. Official Papers of Alfred Marshall: A Supplement.
Cambridge: Cambridge University Press.
26
Marshall, Alfred. 1890. “Some Aspects of Competition.” Reprinted in A.C. Pigou, ed.,
Memorials of Alfred Marshall. London: Macmillan, 1925, pp. 256-91.
______. 1907. “Social Possibilities of Economic Chivalry.” Reprinted in A.C. Pigou, ed.,
Memorials of Alfred Marshall. London: Macmillan, 1925, pp. 323-46.
______. 1917. “National Taxation after the War.” In William Harbutt Dawson, ed., AfterWar Problems, by The Earl of Cromer et al. New York: Macmillan.
______. 1923a. Industry and Trade, 4th edn. London: Macmillan.
______. 1923b. Money, Credit and Commerce. London: Macmillan.
______. 1926. “Memorandum on fiscal policy of international trade.” In J.M. Keynes,
ed., Official Papers by Alfred Marshall. London: Macmillan.
Medema, Steven G. 2005. “‘Marginalizing Government’: From La Scienza delle Finanze
to Wicksell.” History of Political Economy 37 (Spring): 1-25.
______. 2007a. “The Hesitant Hand: Mill, Sidgwick, and the Evolution of the Theory of
Market Failure.” History of Political Economy 39 (Fall): 331-58.
_______. 2007b. “Sidgwick's Utilitarian Analysis of Law: A Bridge from Bentham to
Becker?” American Law and Economics Review 9 (Spring): 30-47.
______. 2009a. The Hesitant Hand: Taming Self-Interest in the History of Economics
Ideas. Princeton: Princeton University Press.
_____. 2009b. “Adam Smith and the Chicago School.” In Jeffrey Young, ed., The Elgar
Companion to Adam Smith, Aldershot: Edward Elgar Publishing, forthcoming.
Mishan, E.J. 1971. “The Postwar Literature on Externalities: An Interpretive Essay.”
Journal of Economic Literature 9 (March): 1-28.
Municipal and Private Operation of Public Utilities. Report to the National Civic
Federation. 3 vols. New York, 1907.
27
Peart, Sandra J. and David M. Levy. 2005. The Vanity of the Philosopher: From Equality
to Hierarchy in Post-Classical Economics. Ann Arbor, MI: University of
Michigan Press.
Peltzman, Sam. 1976. “Toward a More General Theory of Regulation.” Journal of Law
and Economics 19 (August): 211-40.
Pigou, A.C. 1912. Wealth and Welfare. London: Macmillan.
______, ed. 1925. Memorials of Alfred Marshall. London: Macmillan.
______. 1932. The Economics of Welfare, 4th edn. London: Macmillan.
______. 1935. “State Action and Laisser-Faire.” In Economics in Practice: Six Lectures
on Current Issues. London: Macmillan.
Sidgwick, Henry. (1886) “Economic Socialism.” In Miscellaneous Essays and Addresses.
London: Macmillan, 1904.
______. 1897. The Elements of Politics, 3rd edn. London: Macmillan.
______. 1901. The Principles of Political Economy, 3rd edn. London: Macmillan.
Smith, Adam. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations,
Oxford: Oxford University Press, 1976.
Stigler, George J. 1971a. “The Theory of Economic Regulation.” Bell Journal of
Economics and Management Science 2 (Spring): 137-46.
______. 1971b. “Smith’s Travels on the Ship of State,” History of Political Economy 3
(Fall): 265-77. Reprinted in The Economist as Preacher and Other Essays,
Chicago: University of Chicago Press, 1982.
Wicksell, Knut. 1896. Finanztheoretische Untersuchungen, Jena: Gustav Fischer.
28